ABSTRACT Glioblastoma (GBM) is among the most lethal human malignancies. Median survival is a dismal 12-15 months in spite of aggressive treatment with surgery, radiation and concurrent temozolomide. New avenues are urgently needed for development of effective therapeutics. An emerging area of interest in cancer is the role of the mitochondria, cellular organelles which provide most of the energy needs of the body via oxidative phosphorylation (OXPHOS). Mitochondria possess an independent circular 16.5KB genome (mtDNA) that is an essential, and often overlooked, cytoplasmic element of the human genome. mtDNA encodes proteins critical to OXPHOS and is highly polymorphic in human populations and also subject to somatic variation. A growing body of evidence indicates that ancient and recent germline mtDNA variants and the burden of somatic alterations in mtDNA may play important roles in cancer risk and progression. In glioma, recent evidence points to OXPHOS enzyme activity and tissue mtDNA content playing a role in drug resistance and tumor grade. In this R21, we propose to carry out the first comprehensive investigation of germline and somatic mtDNA variation in GBM survival. The project is based on a unique resource of germline and tumor DNA from large numbers of GBM patients for whom uniform tumor and treatment data, and vital status are available. Germline mtDNA sequence will be determined using high-throughput next-generation sequencing of oral DNA samples. We will investigate whether germline variants distinguish infrequent `extreme survivors' (ES) of 3 years or longer from average survivors (AS) of 15 months or less (150 ES versus 450 AS matched for age and treatment). Mitochondrial haplogroups, common individual variants, as well as rare and singleton variants considered in aggregate within genes and functional pathways, will all be examined for association with patient outcome. We will then investigate, in exploratory analysis, whether somatic variants in tumor mtDNA contribute to patient survival (120 GBM patients with paired germline and fresh frozen primary tumor). Somatic variants will be determined by sequencing native tumor mtDNA employing high depth of coverage to capture low frequency variants within individuals. Structural and single nucleotide variants will be considered by class, total burden and predicted pathogenicity for association with patient outcome. The proposal is responsive to ?Provocative Question 5? from the NCI eg: ? How does mitochondrial heterogeneity influence tumorigenesis or progression?? (http://grants.nih.gov/grants/guide/rfa-files/RFA-CA-15-009.html). Results will contribute new insight on questions surrounding the role of the mitochondrial genome in driving GBM and may point to novel therapeutic strategies that target mitochondrial function.